1
AMR Biosurveillance System in the Italian
Region of Veneto
Session 311, February 15, 2019
Arianna Cocchiglia, R&D Manager, Arsenàl.It Consortium,
Lara Tramontan, Data Manager, Arsenàl.It Consortium
2
Arianna Cocchiglia, R&D manager
Lara Tramontan, Data manager
Have no real or apparent conflicts of interest to
report.
Conflict of Interest
3
• Who we are: Arsenàl.IT and Veneto Region
• Background: the AMR (Antimicrobial
resistance) threat and the need of a surveillance
system
• Our ecosystem: from “islands” to HIE
• Second usage of data: our automatic and real
time updated surveillance system
• A happy end: a working AMR surveillance
system
• Our learned lessons..
Agenda
4
• Describe how standardization, interoperability, sharing and engagement contribute
to build a real time biosurveillance system
• Illustrate the advantages of the adoption of an integrated common semantic in
order to easily interpret and share findings about international public health issues
as antimicrobial resistance
• Explain that monitoring processes can be the foundation of a biosurveillance
integrated system in which many structures collaborate to share information
• Identify the true potential of health data standardization in the context of our
globalized society: sharing information in order to unify procedures all across the
world and to create a common heritage, maintaining reliability in each level
• Recognize how existing information can find a second usage, shifting from the
singular identity to a global vision and considering data in their totality
Learning Objectives
5
Arsenàl.IT - Who we are
6
Where we are
Venice
about 5 million people
7
The Veneto Region
1 regional authority
❖ 7 provinces (counties)
➢ 575 municipalities
1 regional public health system
❖ 9 local health authorities (LHAs)
❖ 2 hospital trusts (HTs)
❖ 1 oncology research center
The Healthcare System
Administrative Framework
8
The AMR phenomenon
Antimicrobials are essential for
infection treatments.
Antimicrobial resistance (AMR)
is the ability of a microbe to resist
the effects of medication that
once could successfully treat the
microbe.
AMR is a natural phenomenon
intensified by repeated exposure
to antimicrobials treatments.
In the last decades, the misuse
and abuse of these drugs
accelerated this process.
➢ carbapenems: a serious threat!!
last treatment option for patients infected with resistant
bacteria!
9
The AMR: a global & local threat
❖ CDC: “more than 2 million illnesses and about 23,000 deaths in the
United States annually”
❖ ECDC: “about 25,000 deaths in Europe annually”
Italy: some of the highest levels of
resistance in Europe!!
➢ a resistance to carbapenems of
33.5% (31-36 95%CI) vs an
european average of 8.1% (8-8
95%CI) for Klebsiella
Pneumoniae (2015)
➢ a resistance to third-generation
cephalosporins of 30.1% (29-31
95%CI) vs an european
average of 13.1% (13-13
95%CI) for Escherichia Coli
(2015)
10
Some suggestions from CDC..
How to combat AMR
11
● Global vision (microbiology data from around the world) but still
limited number of sending structures involved (47 reporting
laboratories in Italy, only 2 belonging to Veneto Region)
● Data reported frequently, but with still a delay of some months
AMR surveillance systems
WHO ECDC
Main international initiatives..
12
• timely provision of information (resistance pattern may vary
over time)
• wide geographical coverage (resistance pattern vary in
location)
• high data completeness and reliability
• available linked information about the patient’s attributes (e.g.
age, gender, clinical condition)
➢ to inform decision making and action!
➢ public health: new policies and guidelines
➢ clinicians: right and prompt empirical treatments
Surveillance systems requirements
13
❖ Emergency needs: e.g.
➢ Veneto Region: death rate for septicemia is 16.5 per
100,000 inhabitants (2013-2016)
➢ in 2017 and 2018 resistant Klebsiella pneumoniae caused
many deaths
❖ Italian’s National Institute of Health launched a National
AMR monitoring program: a network of labs located all over the
territory has been required
❖ Public health and clinicians needs: they started to ask for an
AMR surveillance system
Why an AMR surveillance system
in Veneto?
14
In 2012, there were 23 different
realities..
15
Same roads and same rules, same STANDARD
infrastructure..
➢ they started DOCUMENT SHARING
They started to get in touch..
16
Same SYNTAX, same STANDARD document..
They started to talk..
17
Same STANDARD LANGUAGE
They started to understand each
other..
18
Veneto HIE System
19
Secondary usage of data
Trusted Third Party
(Veneto Region)
HIE system
Italian HIE
systems have, by
law, 3 purposes:
1. Care
2. Governance
3. Research
Secondary data
usage
20
A standard secondary data usage
infrastructure
❖ Automatic and real time data
collection
❖ Standard queries to retrieve
clinical documents of interest
(either “push” and “pull” options)
❖ ETL process for extracting data
from documents (thanks to
standard document syntax and
semantic)
❖ automatic and real time
anonymization process
❖ pseudonymization process
allowed -> longitudinal
analysis and linkage of
different kinds of data
21
Community
The AMR surveillance infrastructure
Hospital
Hospital
EHR (XDS
Doc Source
&
Consumer)
Laboratory
(XDS Repository)
Lab
report
LIS
(XDS Doc
Source)
Regional Information System
XDS Doc
Registry
ePrescriptions &
eReferrals
(XDS
Repository)
GP’s EHR
(XDS Doc
Source &
Consumer)
ePrescription
(e.g. antibiotics)
eRefferal
(e.g. antiobiogram
prescription)
ePrescri
ption
eRefferal
TTP
(anonymization)
Web-based tool
for AMR
monitoring
Anonymous real
time updated DB
● ePrescriptions
● eReferrals
● Microbiology
reports
● Discharge
Letters..
22
Our AMR surveillance system
Real time data availability
Wide geographical coverage: all LHAs in
Veneto Region (except one which will be soon
integrated), both for outpatient and inpatient
care setting
high data completeness: LHAs are
monitored every day about their data
completeness (> 90%, otherwise they
lose money!)
high data reliability: all LISs
certified through a testing
process (pre-production
tests) and quality indicators
monitored automatically
every day (post-production
assessment)
Data linkage: possibility to stratify
the analysis for many patient’s
demographic and clinical
characteristics (e.g. age, gender,
clinical condition)
Value of data made
available thanks to an
interactive & user
friendly web-based
AMR monitoring tool
23
The AMR tool is still in a prototype version.
It is composed of 3 dashboards:
➢ Bacteria isolation dashboard
➢ Bacterial resistance dashboard
➢ Antibiotic usage dashboard
The web-based AMR tool
24
Bacteria isolation dashboard
25
Bacteria isolation dashboard
26
Bacteria isolation dashboard (sepsis)
27
Bacteria isolation dashboard (sepsis)
28
Bacteria isolation dashboard
Different geographical pattern
LHA a
LHA b
29
Bacteria isolation dashboard
Different age range
Pediatric population
Adult population
30
Bacterial resistance dashboard
Klebsiella pneumoniae: AMR profile (% of resistant isolated bacteria stratified by antibiotic type)
Cephalosporins,
4th Generation
31
Klebsiella pneumoniae: AMR profile (% of resistant isolated bacteria stratified by antibiotic type)
Bacterial resistance dashboard
32
Klebsiella pneumoniae: AMR profile (% of resistant isolated bacteria stratified by antibiotic type)
Bacterial resistance dashboard
33
Bacterial resistance dashboard
LHA a
Veneto Region
LHA b
Escherichia coli: AMR profile (% of resistant isolated bacteria stratified by antibiotic type)
• 20,2% of isolates R to
aminoglycosides (tobramycin
& gentamicin)
• 30,1% of isolates R to
cephalosporins, 3rd Gen.
reported by EARS-net for Italy
34
Bacterial resistance dashboard
Adult
population
(>=65)
Pediatric
population
(0-18)
Escherichia coli: AMR profile (% of resistant isolated bacteria stratified by antibiotic type)
35
Antibiotic usage dashboard
36
Antibiotic usage dashboard
37
The AMR tool is still in a prototype version, we are defining it in a working
group with also clinicians, researchers and Regional public health officers
The final users will be both clinicians (GPs and hospital doctors) and public
health officers, moreover it will be made available for research
This tool is aimed to:
➢ support clinicians as a Decision Support System for empiric
prescription
➢ allow public health officers to monitor the AMR status in Veneto
Region, in hospitals and in the community
➢ allow public health officers, researchers and clinicians to define new
and updated guidelines for appropriate antibiotic prescription
➢ we will to integrate the patient’s AMR status in the Regional drug-drug
interaction clinical support system
Progress of work
38
Main issue: Resistance to change!!!
Adoption of new standards (new vocabularies & new document
structure) for laboratory staff: a really hard task
➢ A “mandatory” request to change is important (a change
imposed from the top)
but not enough: the real change come from the bottom
➢ meetings have been organized to engage and motivate
➢ the advantages of standard have to be shown (precision,
accuracy: e.g. only one LOINC® code convey many
precise information: sample, method, metrics)
➢ laboratories staff was assisted by a specifically-developed
system for translating codes
It was not an easy journey..
39
Some technical issues..
❖ Choice of public, international or at least national codifications,
as much complete as possible
➢ a germ codification that meets the requested criteria was not found -> new
codify is being created by the Italian national association of clinical
microbiologists
➢ Some codes have been added to HL7 ObservationMethod and HL7
SpecimenType to include specific Veneto Region codes
❖ Laboratory and microbiological CDA2 reports have a very
complex structure!
➢ High quality reached trough:
i. Labeling process (pre-production tests)
ii. Daily automatic evaluation of quality indicators (in production stage)
➢ Time consuming ETL process:
i. a cloud solution & many parallel process can help!
It was not an easy journey..
40
The creation of a real time and evidence-based biosurveillance system is not
a local, but a Regional/National asset
The drivers:
★ Sharing
★ Standardization
★ Integration
★ Interoperability
But the change is not (only) a technical issue!
Standardization does not reduce information, it can be an enrichment,
because it encourages the sharing of the local knowledge and the
learning between people
Some keynotes for you..
